A Virtual Agent to Support Individuals Living With Physical and Mental Comorbidities: Co-Design and Acceptability...

Area of Science:

• Digital Health
• Artificial Intelligence in Healthcare
• Patient Self-Management

Background:

• Co-occurring mental health issues are common in individuals with long-term physical health conditions, impacting distress and health outcomes.
• Digital tools, including AI-powered virtual agents, are increasingly promoted for patient self-management as healthcare shifts to community settings.
• Autonomous virtual agents utilize AI to understand and respond to natural language, offering a novel approach to health support.

Purpose of the Study:

• To co-design the content, functionality, and interface of an AI virtual agent for chronic pulmonary obstructive disease (COPD) self-management.
• To assess the acceptability and system content of the developed virtual agent with patients and healthcare professionals.

Main Methods:

• Conducted two co-design workshops and a proof-of-concept implementation of an AI virtual agent.
• Utilized natural language processing capabilities for the virtual agent.
• Performed video-based scenario testing for acceptability with adults diagnosed with COPD and their healthcare providers.

Main Results:

• Identified four priority self-management scenarios: diagnosis information, exacerbation crisis support, low mood emotional support, and general motivation.
• 12 additional adults with COPD found the system acceptable and engaging, especially with Internet of Things (IoT) integration.
• The virtual agent was perceived as particularly useful for individuals living alone.

Conclusions:

• Patients do not compartmentalize mental and physical health needs; the virtual agent content reflected a psychological focus.
• Supported self-management via an autonomous virtual agent was deemed acceptable by participants.
• Co-design identified key principles for developing AI agents to support self-management in COPD and other long-term conditions.